Optimization of optimal connection mode of high voltage distribution network based on LSTM and random matrix theory

The wiring pattern of high-voltage distribution network is disturbed by inductive load and harmonic load, which affect the current and voltage waveforms in the power network, causing the power factor to deviate from the ideal state, resulting in a large power difference, resulting in poor optimization effect of the wiring pattern of high-voltage distribution network. Therefore, this paper puts forward the optimization of the optimal connection mode of high voltage distribution network based on LSTM and random matrix theory. In this paper, LSTM is introduced, and the characteristics of connection mode are deeply weighted and learned. The optimal topology identification of high-voltage distribution network is realized through the weighted fusion characteristics of forgetting gate, memory gate and output gate mechanism. Combining the random matrix theory, the analysis ability of the model for topological complexity is enhanced. Taking load change, voltage stability and equipment status as optimization objectives, the power difference affected by wiring mode is analyzed to realize the optimal wiring mode optimization. The experimental results show that the power difference is basically zero and the power factor is close to the ideal state after applying this method. This method can accurately optimize the power supply and the optimization effect is better.